Advocacy Learning: Learning through Competition and Class-Conditional Representations
| Autor: | Ian Fox, Jenna Wiens |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Scheme (programming language) Computer Science - Machine Learning Computer science business.industry Computer Vision and Pattern Recognition (cs.CV) Supervised learning Computer Science - Computer Vision and Pattern Recognition Machine Learning (stat.ML) Class (biology) Machine Learning (cs.LG) Competition (economics) Discriminative model Statistics - Machine Learning Argument Artificial intelligence Baseline (configuration management) business Representation (mathematics) computer computer.programming_language |
| Zdroj: | IJCAI |
| Popis: | We introduce advocacy learning, a novel supervised training scheme for attention-based classification problems. Advocacy learning relies on a framework consisting of two connected networks: 1) $N$ Advocates (one for each class), each of which outputs an argument in the form of an attention map over the input, and 2) a Judge, which predicts the class label based on these arguments. Each Advocate produces a class-conditional representation with the goal of convincing the Judge that the input example belongs to their class, even when the input belongs to a different class. Applied to several different classification tasks, we show that advocacy learning can lead to small improvements in classification accuracy over an identical supervised baseline. Though a series of follow-up experiments, we analyze when and how such class-conditional representations improve discriminative performance. Though somewhat counter-intuitive, a framework in which subnetworks are trained to competitively provide evidence in support of their class shows promise, in many cases performing on par with standard learning approaches. This provides a foundation for further exploration into competition and class-conditional representations in supervised learning. Comment: Accepted IJCAI 2019 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|